As evidenced by Healthy People 2010's goal of eliminating health disparities, the large and persistent disparities in health in the United States (US) are an area of tremendous concern. Identifying health disparities and understanding their root causes are essential prerequisites for developing appropriate policy recommendations and designing effective interventions to address and prevent disparities. This proposal is designed to test a fundamental cause theory-driven approach to examine in detail the production of disparities in mortality over time in the US from HIV/AIDS and lung cancer. Fundamental cause theory holds that as the capacity to control disease and prevent death increases, this expanded capacity combines with existing social and economic inequalities to create health disparities by race and socioeconomic status (SES). Therefore, when new treatments, as in the case of HIV/AIDS, or new prevention strategies, as in the case of lung cancer, become available, it is predicted that these developments will disproportionately benefit those with better access to resources and therefore result in disparities. In contrast, the theory predicts no significant changes in disparities over time for diseases for which there is no new information or treatment. To test this theory, case studies of the following three categories of disease will be analyzed in detail: 1) HIV/AIDS, an infectious disease for which there have been major advances in treating disease and delay death, 2) lung cancer, a disease for which there have been major improvements in knowledge of how to prevent disease, and 3) pancreatic cancer, a disease for which there have been no major changes in prevention or treatment. Mortality data from the National Center for Health Statistics from 1968 to 2004 and county level census data from the same time period will be used to construct all-cause and cause-specific age-, sex-, and race- specific mortality rates for HIV/AIDS, lung cancer, pancreatic cancer for every county in the United States. County-level SES measures will be constructed for the decennial census from 1960 to 2000. These sources of data will be merged into a single data base and prepared for analysis. This project will generate several products to enhance public health: 1) the results from this study will help to further elucidate the causes of the development of disparities, 2) a series of policy recommendations will be developed based on the findings, and 3) the methods used to create the data base will be made available to other investigators. [unreadable] [unreadable] [unreadable] [unreadable]